Determining an appropriate control for use in acupuncture research remains one of the largest methodological challenges acupuncture researchers face. In general, acupuncture controls fall under one of two categories: (1) sham acupuncture, in which the skin is punctured with real acupuncture needles either fully at nonacupoint locations or shallowly at acupoint locations or both and (2) placebo acupuncture, which utilizes nonpenetrating acupuncture devices. In this study, we will focus on non-penetrating placebo acupuncture devices (blunted-needle and nonneedle devices) that are currently available in acupuncture research. We will describe each device and discuss each device’s validation and application in previous studies. In addition, we will outline the advantages and disadvantages of these devices and highlight how the differences among placebo devices can be used to isolate distinct components of acupuncture treatment and investigate their effects. We would like to emphasize that there is no single placebo device that can serve as the best control for all acupuncture studies; the choice of an acupuncture control should be determined by the specific aim of the study.
Over the past decades, acupuncture treatment has gained popularity in the Western world due to its therapeutic effect. However, studies have achieved contradictory results when using control treatments to test the true efficacy of acupuncture. Studies consistently show that both real and placebo acupuncture treatments confer significant benefits over no-treatment control conditions [
Placebo research has revealed several important variables related to acupuncture treatment that can be either modulated or held constant between verum (real) and placebo (control) conditions in order to test the effects of specific and nonspecific components of acupuncture. In general, acupuncture involves the insertion of needles into the body; thus, one of the components of verum acupuncture is skin penetration. While this particular component is not held constant between verum and non-penetrating acupuncture controls, there are several other variables that can be either held constant or modulated, which include sensory stimulation, dose (number of needles), acupoint location, practitioner-patient interaction, and treatment setting.
In acupuncture research, double-blinded randomized clinical trials (RCTs) serve as the gold standard when comparing the effects of a specific treatment to the effects of a placebo control. In acupuncture RCTs, it is ideal for the control treatment to be both physiologically inert and indistinguishable from the real treatment. Thus, an effective inert treatment in the control condition is essential in order for a double-blinded RCT to achieve a high level of scientific validity. Determining the proper inert control for an RCT designed to evaluate the efficacy of acupuncture is methodologically challenging for three main reasons. It is difficult to create an inert control device that mimics both the visual appearance of the acupuncture treatment device and the method of needle insertion involved in acupuncture treatment. It is challenging to develop an inert acupuncture device that can control for all nonspecific factors involved in an acupuncture treatment. The therapeutic effect of acupuncture relies on several nonspecific factors, including the ritual procedure of acupuncture administration, the patient-practitioner interaction, the nature of the illness, the treatment, and the treatment setting [ It is challenging to make the acupuncturist who is directly administrating the acupuncture blind to treatment condition.
As a result, over the years, investigators have used a variety of controls in their studies to account for these challenges in terms of their own specific study aims. In acupuncture research, placebo controls for acupuncture studies fall under one of two categories: (1) sham acupuncture, in which the skin is punctured with real acupuncture needles either fully at nonacupoint locations or shallowly at acupoint locations, or both and (2) placebo acupuncture, which utilizes non-penetrating acupuncture devices. In this review, we will focus only on non-penetrating placebo acupuncture devices, including blunted needle and nonneedle devices that are currently available for use in acupuncture research.
Subsequently, the Streitberger device has been used as a control in acupuncture research [
Unlike the Streitberger device, the Park device includes a Park tube that holds the placebo needle more firmly in place and facilitates lifting and thrusting, a common manual stimulation technique in acupuncture. Due to the fact that the guide tube is fixed to the skin, the Park device cannot be used to insert the needle at certain angles or depths. This limitation is problematic when using acupoints requiring shallow or horizontal insertion.
Since the original validation studies, the Park device has been used in several acupuncture RCTs as a placebo acupuncture control [
In a study of 36 rheumatoid arthritis patients conducted by Tam and colleagues [
In another validation study using a crossover design [
Plastic guide tubes are also used as placebo devices, as they provide the sensation of acupuncture stimulation but do not penetrate the skin [
Trials involving these devices are difficult to replicate because the devices are not standardized. Since the majority of these trials have not been precisely replicated (i.e., the devices are slightly different), the validity of the results of RCTs that use these devices remains to be tested.
A new variation of TENS device worth noting is a special electrode designed for ear stimulation. It consists of two pair carbon-impregnated silicone electrodes fixed to one ear clamp; only one pair of the electrodes is connected to the electrical wire. Given that the electrode wiring is imbedded in the clamp, this design allows for subject and practitioner blinding [
Several subsequent clinical trials have employed placebo TENS to investigate the efficacy of verum TENS. In a study of chronic back pain, a 2 × 2 factorial design was used to compare TENS, placebo TENS, exercise, and no exercise. No superior benefit was found for verum TENS over placebo TENS using a VAS of pain ratings and other clinical outcome measures [
In another study, verum TENS versus placebo TENS was studied in patients with multiple sclerosis and chronic low back pain. After correcting for multiple comparisons, there were no significant differences on pain visual analog scale and other self-evaluation scales between the verum and placebo TENS groups [
In one study, investigators studied the efficacy of TENS compared to placebo TENS and a control no-treatment group in a population of patients with chronic low back pain. Average pain ratings collected immediately before and after each treatment demonstrated a significant reduction in pain in both the verum and placebo TENS groups. Pain intensity was reduced significantly more for those in the TENS group compared to the placebo TENS group. Additionally, investigators studied the additive effects of 10-week treatment (administered twice weekly) and found that TENS but not placebo TENS demonstrated a reduction in pain intensity over the first 16 treatments. Similarly, verum TENS was more effective than placebo TENS in maintaining pain reduction one week after the last treatment. The benefit in pain reduction continued for 3 and 6 months after completion of the study regardless of whether subjects received verum or placebo TENS treatment. The no-treatment control condition, in contrast, demonstrated no natural improvement in pain over the same course of time [
The validity of the placebo laser acupuncture as a general acupuncture control was thoroughly discussed in an article written by Irnich et al. [
Finding a proper control remains a primary methodological concern in acupuncture research. Traditionally, two main categories of acupuncture control have existed: (1) sham acupuncture, which involves the use of real needles that puncture the skin and (2) placebo acupuncture, which is non-penetrating. Sham acupuncture utilizes penetrating needles that are either applied fully to nonacupoints or shallowly to acupoints or both [
Thus, a placebo acupuncture device may be a more appropriate control for verum acupuncture because it minimizes the physiologic response and is relatively inert. In this paper, we have described a range of placebo acupuncture devices currently used in acupuncture research. Table
Considerations for use of placebo devices in acupuncture research.
Types of device | Considerations | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Visually indistinguishable | Somato-sensation | Constraint of needle insertion | Commercial availability | Self-made | Modality specific | Price | Reusable | Appropriate for double-blind study design | Appropriate for crossover study design | |
Blunted needles devices | ||||||||||
The Steitberger |
|
|
|
$6.3 per needle | ||||||
The Park |
|
|
|
|
$2.9 per needle | |||||
The Japanese |
|
|
|
NA |
|
|||||
Foam device |
|
|
|
NA |
|
|||||
Others |
|
|
NA |
|
||||||
| ||||||||||
Nonneedle devices | ||||||||||
Placebo TENS |
|
NA |
|
|
$59 to $260 per unit |
|
|
|
||
Placebo laser |
|
|
NA |
|
|
$hundred to thousand per unit |
|
|
|
Prices according to Google Shopping. The actual prices may vary across vendors.
Placebo acupuncture is generally noninvasive. The blunted needles used in these devices are relatively inert and indistinguishable from real acupuncture needles. The Streitberger, Park, and Japanese devices are three of the most commonly used placebo needle devices in acupuncture research. The Japanese device is the only needling device that can be employed in double-blinded trials. All three of these devices are used in a similar manner; a blunted needle is inserted, touches the skin, and is retracted. These three devices, while standardized and validated, are not widely available for researchers worldwide since they are custom fabricated. Aside from the Streitberger device, all other placebo devices have physical limitations with regard to their ability to facilitate a range of manipulation methods, such as shallow or horizontal insertion. They are also not effective for all acupoint types and acupuncture positions. The foam device, simple toothpick, and blunted needle device are not standardized and thus cannot be validated. Since these devices have not been validated, the quality of the results from RCTs using these devices remains unconfirmed.
All placebo needle devices involve contact with the skin. Physical contact with the surface of the skin may provide sensory stimulation, indicating that even blunted needle devices are not entirely inert and may elicit a therapeutic effect of their own. For instance, Han and Lund et al. noted that the sensation produced by a needle tip “touch” was substantial [
The response to placebo acupuncture treatment is due to a variety of factors, including the presence or absence of deqi sensations. The deqi phenomenon is a complex set of physiological sensations associated with acupuncture treatment [
Similar to other longitudinal treatment studies involving placebo treatments, maintaining patient’s blinding throughout a treatment study involving multiple acupuncture treatments can be difficult. All blunted needle placebo acupuncture devices can be useful in short-term, acute intervention trials. However, in conditions requiring long-term treatment, maintaining patient’s blinding may be difficult due to the subjects’ natural curiosity and/or motivation to learn more about the type of treatment they have been receiving, which can lead to unblinding (e.g., patients may talk to each other, read about acupuncture, or go to another acupuncturist). It is important for investigators to assess whether or not the subject blinding can be maintained through the end of the study.
Unlike verum acupuncture, some placebo devices require tape or foam for successful application of the placebo treatment. Devices that require tape or foam may induce allergic reactions and thus may not be tolerated by all subjects. In addition, there is risk of infection if the needle were to be inserted through the tape and thus it is necessary to use sterile techniques. Additionally, these devices may not be suitable for all acupoints, such as points on the scalp, fingers, and toes that cannot provide a flat surface for tape or foam.
“Nonneedle” placebo acupuncture devices (TENS and laser acupuncture) have their own set of common considerations. Neither placebo TENS nor laser acupuncture provide repeated needle stimulation of the skin and thus are relatively physiologically inert. Nonneedle placebo devices are also effective for both acupuncture-experienced and acupuncture-naive subjects. The major disadvantage of nonneedle devices like TENS and laser acupuncture is that they differ from verum needle acupuncture in design and concept, as well as in context and culture. Therefore, placebo TENS and placebo laser acupuncture may only be considered to serve as valid placebo controls in TENS and laser acupuncture studies, respectively.
Finally, we would like to emphasize that this paper is developed to aid investigators in designing studies that test and explore the efficacy and mechanism of acupuncture and to facilitate the selection of appropriate acupuncture placebo devices. The most appropriate placebo acupuncture devices are those that are the most indistinguishable and inert in consideration of the specific design on the study.
The authors thank the valuable suggestions, comments, and English editing form Sonya Freeman, Rosa Spaeth, and Lisette Roman in the Department of Psychiatry, Massachusetts General Hospital. This work was supported by the National Science Foundation of China (Grant no. 81072768), New Century Excellent Talents in University of Ministry of Education of China (Grant no. NCET-11-0603), Chinese Medicine Clinical Evaluation Research Team in Brain Disease of Beijing University of Chinese Medicine (Grant no. 2011-CXTD-22), R01AT006364, R21AT004497, and R03AT218317 to Jian Kong, and P01AT006663 to Bruce Rosen.